As I write this column shortly after George W. Bush'S inauguration, I find a continuous chorus of dismay about the outcome of the presidential election. Many people argue that it's about accurate vote counts, an issue that engineers at Unisys and Dell and elsewhere are already preparing to address. But as I see it, they are attacking only part of the problem.
Better voting technology means that we will get quicker, more accurate counts. But it will do nothing to address what happens when the results are "too close to call." None of us back in civics class were ever told what was supposed to happen in the case of a statistical dead heat. Only when the votes came up close did we find out that the decision-making process was seriously lacking. Perhaps that's why so many people are still disaffected.
The statistical dead heat can be an engineer's worst dilemma. When the options are all equally good (or bad, if you prefer), flipping a coin might seem like the best solution. But one lesson we've learned from this election is that the process of picking a winnerómatters. A lot.
Since every circumstance is different, there's no "one-process-fits-all" solution. But here's my litmus test to determine whether a process is effective: Everyone may not agree with the ultimate decision, but they nevertheless feel comfortable with it.
Consider your current methods for resolving your own "too-close-to-call" situations. Do they incorporate the following features of a successful process?
The process is thorough.
It is open and accessible to all participants.
Participants know that their input into the process was considered.
The outcome is consistent with all of the above.
If so, congratulations. Of course, just because you have a good tie-breaking mechanism doesn't mean everyone will be satisfied with the outcome. But you will increase the probability they will support the winning solution.
If only politics worked the same way.